What is the consequence of having an error term that jumps at cutoff in a RD model?
A) B1, the coefficient for the treatment effect is biased.
B) B2, the coefficient for the slope is biased.
C) We need to use a polynomial model.
D) We need to use a varying slopes model.

Question

Quizplus · Accepted Answer

An error term that jumps at cutoff in a RD model creates bias in the coefficient for the treatment effect (B1). This is because the error term affects the assignment to treatment and control groups at the cutoff, which in turn affects the estimation of the treatment effect. Therefore, choice A is the best answer.

What Is the Consequence of Having an Error Term That